UNIVERSITA’ DEGLI STUDI DI VERONA
FACOLTA’ DI MEDICINA E CHIRURGIA
Dipartimento di Patologia e Diagnostica
SCUOLA DI DOTTORATO DI
Scienze Biomediche Traslazionali
DOTTORATO DI RICERCA IN
Biomedicina Traslazionale
CICLO XXIV
TITOLO DELLA TESI DI DOTTORATO
APPETITIVE MEMORIES AND RELAPSE TO DRUG-USE: INVESTIGATIONS ON EFFECTS OF SELECTIVE DISRUPTION OF MEMORY MECHANISMS IN A RAT MODEL OF NICOTINE DEPENDENCE
S.S.D BIO14/FARMACOLOGIA
Coordinatore: Chiar.mo Prof. Cristiano Chiamulera Tutor: Chiar.mo Prof. Cristiano Chiamulera
Dottoranda:
Abstract
Tobacco use through cigarette smoking is the leading preventable cause of death in the developed world. The pharmacological effect of nicotine plays a crucial role in tobacco addiction. Nicotine dependence has a huge impact on global health and although several medications are available, including a wide range of nicotine-replacement therapies (NRTs), bupropion, and recently approved nicotinic receptor partial agonist varenicline, at best only about a fifth of smokers are able to maintain long-term (12 months) abstinence with any of these approaches. Thus, there is a need to identify more effective treatment to aid smokers in maintaining long-term abstinence.
Several preclinical and clinical studies have underlined the importance of non-pharmacological factors, such as environmental stimuli, in maintaining smoking behaviour and promoting relapse. Initially neutral stimuli that are repeatedly paired with a reinforcing drug (e.g. lighter) acquire a new conditioned value (conditioned stimuli, CS) and become able to elicit craving even in the absence of the drug. Indeed smokers are particularly reactive to smoking/nicotine related CS, this phenomenon is called cue-reactivity and involves a vast array of physiological, psychological and also behavioural response, such as decrease in heart rate and blood pressure and/or increase in skin conductance and skin temperature, increase in craving and urge to smoke and/or mood change, and also change in smoking behaviour (e.g., latency to smoke, cigarette puff volume and frequency, amount of cigarette consumed and relapse to smoking behaviour). Given the importance of the learned association between stimuli and nicotine in the phenomenon of relapse to nicotine-seeking behaviour, it has been proposed that treatment that disrupts the nicotine-associated memories could act as a pro-abstinent and anti-relapse therapy.
After learning experience, memories are stored by a process called consolidation. For at least a century it has been a dogma that initially labile memory (short-term memory) are consolidated by the passage of time and become stable and permanent (long-term memory). However converging evidence from animal and human studies have revealed that memories may return to a vulnerable phase during which they can be updated, maintained and even disrupted. The retrieval of memory indeed may destabilize the consolidated memories that require a new process to be maintained. This hypothetical process is called reconsolidation. The disruption of drug-related memories reconsolidation has been proposed as a potential therapeutic target to prevent the
CS-3
induced relapse in ex drug-addicts. Several animal studies have shown that the reconsolidation of drug-related memories can be disrupted by the administration of an amnestic drug contingently upon retrieval of the memory. Unfortunately most of the compound used in animal studies has serious tolerability and safety issues in humans. Recently Monfils at al and Schiller et al have shown that it is possible to disrupt fear memory reconsolidation and consequently prevent the return of fear by providing CS-extinction training shortly after retrieval of the memory. CS-CS-extinction consists in the repeated presentations of CS (e.g. lighter) in the absence of the unconditioned stimulus US (e.g. drug) leading to a decrease of the previously acquired conditioned response (e.g. smoking behaviour).
The main objective of the present thesis was to investigated whether it is possible to disrupt nicotine related memories reconsolidation by applying CS-extinction after the retrieval of such as memories, and whether this disruption prevent the relapse to nicotine-seeking behaviour in a rat model of nicotine dependence. Furthermore we investigated also whether nicotine-related memories reconsolidation might be pharmacologically disrupted by administering a drug, that have been shown to disrupt memory reconsolidation in previous literature studies (i.e. propranolol or MK-801), at memory retrieval.
The experimental approach used to address this issue was the paradigm of nicotine self-administration in rats, a paradigm based on Pavlovian and operant conditioning to nicotine and nicotine-associated cues. We performed five experiments in which CS-extinction or the pharmacological treatment (i.e. propranolol or MK-801) was associated to different memory retrieval protocols. We therefore assessed the effect of these post-retrieval treatments on relapse to nicotine or food seeking behaviour. Retrieval consists in presenting the CS in the absence of US, a procedure similar to CS- extinction. Since the length of CS exposure (i.e. number of CS presentations) is a crucial factor for reconsolidation or extinction occurrence, different retrieval length (1, 3 or 30 CS presentations) have been presented to retrieve nicotine-related memories. Results showed that CS-extinction applied after a short retrieval (3 CS presentations) reduced the relapse to nicotine seeking behaviour compared to control groups that did not receive CS-extinction, moreover this effect was not observed when CS-extinction was applied without retrieval. These results suggest that the effect of post-retrieval CS-extinction was specifically due to inhibition of nicotine-related memories reconsolidation. To our knowledge, this is the first evidence of post-retrieval
CS-extinction effect on drug-seeking behaviour. Considering that this is an indirect demonstration of the occurrence of memory reconsolidation process, we would also consider these findings as the first evidence of nicotine Pavlovian memory reconsolidation. On the other hand, no effect of MK-801 or propranolol on nicotine seeking behaviour has been observed. These results are in contrast with other literature data, however methodological issues might explain the contrasting results.
More evidences are needed to confirm that the effect of post-retrieval CS-extinction was due to interference of CS-extinction with reconsolidation process. Further studies will investigate the effect of CS-extinction applied 6 hours after retrieval, a delay time that allows to apply CS-extinction outside the labile phase of memory due to retrieval. Moreover it would be fundamental to identify specific molecular markers of reconsolidation or extinction. To find a selective molecular correlate of reconsolidation will allow to disentangle the point of whether our retrieval protocols are inducing reconsolidation or extinction and will provide further evidence that post-retrieval CS-extinction interfere with reconsolidation of CS-memory. This could also be useful to better understand the lack of effect of MK-801 and propranolol in our experiments. It has been pointed out by Lee & Everitt (2008) that to successfully reactivate a memory acquired instrumentally (as in our experiments) the CS should be presented contingently upon acquired response. We can then hypothesized that presenting the CS contingently upon response during retrieval session, would lead to a more strong retrieval and destabilization of the memories, and to a stronger effect of CS-extinction and of MK-801 on the reconsolidation of that memory.
Finally, it would be important to assess whether the effect of post-retrieval CS-extinction on nicotine seeking behaviour is persistent, by repeating the test several week after retrieval-CS-extinction procedure.
In conclusion, our findings suggest that the exposure to nicotine CS-extinction, after a short retrieval of the same nicotine CS, may inhibit CS-induced relapse to nicotine-seeking behaviour and may offer a potential co-adjuvant to current therapeutic interventions for smoking cessation and abstinence maintenance.
5
CONTENTS
1. INTRODUCTION...7
1.1.NEUROBIOLOGY OF NICOTINE...8
1.1.1. Absorption... 8
1.1.2 Nicotinic cholinergic receptors and neuroadaptation... 8
1.1.3 Nicotine and neurotrasmitters release... 9
1.1.4 Nicotine effects and withdrawal...12
1.1.5 Pharmacological smoking cessation treatment...13
1.2PSYCHOBIOLOGY OF TOBACCO ADDICTION...14
1.2.1 Conditioning...14
1.2.2 Nicotine’s multiple-action...15
1.2.3. Cue reactivity...15
1.3NICOTINE-RELATED MEMORIES...16
1.3.1 Reconsolidation theory...16
1.3.2 Reconsolidation as a potential target in drug addiction treatment...21
1.3.3 Extinction therapy...23
1.3.4 Extinction and reconsolidation interaction...23
1.4AIM...28
1.4.1 Experimental model...30
2. MATERIALS AND METHODS... 33
2.2.DRUGS...33
2.3EXPERIMENT 1...34
2.3.1 Apparatus...34
2.3.2. Training to lever press...34
2.3.3. Surgical procedure...34
2.3.4. Training to nicotine self-administration (S/A)...35
2.3.5 Retrieval...35
2.3.6 Treatment...35
2.3.7 Renewal...36
2.4EXPERIMENT #2...36
2.4.1 Apparatus...36
2.4.2. Training to lever press...36
2.4.3. Training to food self-administration (S/A)...36
2.4.4 Retrieval...37
2.4.5 Treatment...37
2.4.6 Renewal...37
2.5.EXPERIMENT #3...37
2.5.1 Apparatus...37
2.5.2. Training to lever press...38
2.5.3. Surgical procedure...38
2.5.4. Training to nicotine self-administration (S/A)...38
2.5.5 Retrieval...39
2.5.6 Treatment...39
2.5.7 Renewal...39
2.6EXPERIMENT #4...40
2.6.1 Apparatus...40
2.6.2. Training to lever press...40
2.6.3. Surgical procedure...40
2.6.4. Training to nicotine self-administration (S/A)...41
2.6.6 CS presentation...41
2.6.7 Renewal...42
2.7.EXPERIMENT #5...42
2.7.1 Apparatus...42
2.7.2. Training to lever press...42
2.7.3. Surgical procedure...42
2.7.4. Training to nicotine self-administration (S/A)...43
2.7.5. Instrumental learning extinction phase (ILEXT)...43
2.7.6 Retrieval...43 2.7.7 Treatments...44 2.7.8 Renewal...44 2.8DATA ANALYSES...44 3. RESULTS... 45 3.1.THE MODEL...45
3.1.1. Food self-administration acquisition....45
3.1.2 Nicotine self-administration acquisition...46
3.1.3 Instrumental learning extinction phase...50
3.1.4 Renewal...51 3.2.THE PROJECT...53 3.2.1. Experiment #1...53 3.2.2. Experiment #2...56 3.2.3 Experiment #3...61 3.2.4. Experiment #4...65 3.2.5 Experiment #5...68 4. DISCUSSION... 74 4.1.CONCLUSION...84 5. REFERENCES... 87
7
1. INTRODUCTION
Tobacco use is the leading global cause of preventable and premature death. It is one of the main causes for a number of chronic diseases, including cancer, lung diseases, and cardiovascular diseases. Cigarette smoking is also a risk factor for respiratory tract infections, reproductive disorders, osteoporosis, adverse postoperative events such as delayed wound healing, duodenal and gastric ulcers and diabetes (Vineis et al, 2004). Tobacco use kills nearly 6 million people and causes hundreds of billions of dollars of economic damage worldwide each year. If the current trends continue, by 2030 tobacco will kill more than 8 billion people worldwide each year (World Health Organization Report 2011). Seventy percent of smokers say that they would like to quit, eighty percent who attempt to quit on their own return to smoking within a month, and each year, only 3% of smokers quit successfully.
Smoking-related diseases are a consequence of prolonged exposure to toxins in tobacco smoke; therefore the most dangerous aspect of smoking is that constituents are highly addictive.
Tobacco addiction is reported both in the Diagnostic and Statistical Manual of Mental Disorders, 4th edn. and in the World Health Organization’s International Classification of Diseases, version 10.
The criteria for defining drug dependence are the following: Primary criteria:
• Highly controller or compulsive use
• Psychoactive effects
• Drug-reinforced behaviour Additional criteria:
• Addictive behaviour often involves -Stereotypic patterns of use
-Use despite harmful effects -Relapse following abstinence -Recurrent drug cravings
• Dependence-producing drugs often produce -Tolerance
-Physical dependence
Tobacco dependence fit all the above criteria. It is a behavioural disorder due to chronic exposure to a psychoactive substance, nicotine (Abrams et al., 1999). Importantly, smokers do not just self-administer nicotine while smoking, but they experience the pharmacological effect of nicotine in a context rich of environmental stimuli. Indeed, tobacco addiction arises from an interplay of i) pharmacological effect of nicotine, ii) psychological and physiological susceptibility of the individual (e.g. genetic predisposition, psychiatric disorder, impulsivity) and iii) social and environmental influences (including tobacco product and marketing) (Caggiula et al., 2001, Field et al., 2009; Karp et al., 2006; Pomerlau, 1995; Rodriguez et al., 2007).
1.1. Neurobiology of Nicotine
1.1.1. Absorption
Nicotine is an alkaloid that constitutes approximately 0.6–3.0% of the dry weight of tobacco. It is a psychoactive addictive drug. Inhalation of smoke from a cigarette distils nicotine from the tobacco in the cigarette. Smoke particles carry nicotine into the lungs, where it is rapidly absorbed into the pulmonary venous circulation. The nicotine then enters the arterial circulation, rapidly crosses the blood barrier in approximately 7-10 seconds and move into the brain, where it binds to nicotinic cholinergic receptors (nAChR) (Hukkanen et al., 2005).
1.1.2 Nicotinic cholinergic receptors and neuroadaptation
nAChR is a ligand-gated ion channel that normally binds acethylcholine (Albuquerque et al., 2009). It consists in five peptidic subunits: the mammalian brain expresses nine α subunits and three β subunits. Usually the receptor is composed of two α and three β subunits arranged to form a pore (Jensen et al., 2005). The receptor α4β2 is the most abundant and the principal mediator of nicotine dependence. Ligand binding occurs via the α subunit, producing a conformation change that opens the cationic channel and allow sodium and calcium ion influx, after few milliseconds the channel close and become desensitised. In the absence of agonist, the receptor return to the standby stage where it is closed but “activable”. Moreover chronic nicotine exposure increases nicotine or acethylcholine (ACh) binding in the brain, a phenomenon known as up-regulation.
9
receptors return to the standby state leading to a hyperexcitability of cholinergic system. This hyperexcitability is associated with withdrawal effect: the symptoms of craving and withdrawal, indeed, begin in smokers when the up-regulated desensitized α4β2 receptors become responsive during a long period of abstinence, such as overnight. Nicotine binding of these receptors during smoking alleviates craving and withdrawal (Dani & Heinemann, 1996). Smokers regulate the daily amount of cigarette smoking in order to maintain near-complete saturation, and thus desensitization, of the α4β2 receptors. Thus smokers are probably attempting to avoid withdrawal syndrome when maintaining a desensitized state.
1.1.3 Nicotine and neurotrasmitters release
nAChRs are localized mainly at presynaptic level on a number of different type of neurons, such as on glutamatergic, on dopaminergic, noradrenergic and gamma aminobutyric acid (GABA) neurons in the ventral tegmental area (VTA), substantia nigra, and striatum (Figure 1). Thus nicotine modulates not only ACh level but also dopamine (DA), glutamate and GABA activity (Albuquerque et al., 1997; Alkondon et al., 1997; Gray et al., 1996; Guo et al., 1996; Ji et al., 2001; Jones et al., 1999; Jones & Wonnacot, 2004; Li et al., 1998; Mansvelder & McGehee, 2000; Marubio et al, 2003; McGehee & Role, 1995; McGehee et al., 1995; Radcliffe & Dani, 1988; Radcliffe et al., 1999; Role & Berg, 1996; Wonnacott, 1997, Yin and French, 2000)
Figure 1: Schematic drawing of dopaminergic, gabaergic, glutamatergic and cholinergic neurons interaction. nAChRs are localized mainly at presynaptic level on glutamatergic, dopaminergic, and gabaergic neurons. Abbreviations in the text, except D1 = dopamine receptor 1, mGluR2/3 = metabotropic glutamate receptor type 2or 3. Image taken from Balfour, 1994.
It is widely accepted that nicotine dependence, similarly to other drugs of abuse (such as cocaine, amphetamine, etc.), arises from nicotine action on dopaminergic neurons in the mesocorticolimbic system. This system is also called reward pathway and involves dopaminergic neurons located in VTA and their projection into the striatum, amygdala, prefrontal cortex and the shell of nucleus accumbens (Figure 2).
11
Figure 2: Schematic drawing of mesocorticolimbic pathway, mediating nicotine dependence. Nicotine stimulates nAChR located in the VTA, resulting in release of DA in the nucleus accumbens. Neurons projecting from the prefrontal cortex and amygdala modulate the release of DA in the nucleus accumbens. GABAergic neurons projections modulate DA release in nucleus accumbens and VTA. Image taken from Le Foll & George, 2007.
It has been well established that the activation of mesocorticolimbic DA pathways is associated with drug reward (Di Chiara, 2000), where increased neuronal firing in the VTA (Clarke, 1990; French, et al., 1996) and DA release in the nucleus accumbens (Di Chiara and Imperato, 1988) are neurochemical correlates of psychostimulant self- administration. Laboratory animals self-administer nicotine, indicating that the drug exerts effects on mesocorticolimbic DA neurotransmission in a comparable manner to other psychostimulant drugs of abuse. Supporting a predominant role for enhanced dopaminergic neurotransmission, nicotine concentrations self- administered by rodents and humans also increase DA release in the nucleus accumbens (Imperato, et al., 1986; Nisell, et al., 1994) and activate DA neurons in the VTA (Pidoplichko, et al., 1997).
Moreover it has been shown that inhibition of DA release in nucleus accumbens by antagonist drugs attenuates reinforcing properties of nicotine, leading to a decrease in nicotine self-administration in rats (Corrigal & Coen, 1989; Stolerman & Shoaib, 1991.).
As stated above, nicotine also augments both glutamate and GABA release: the former one facilitates DA release, the latter inhibit DA release. Chronic exposure to nicotine induces desensitization of some types of nAChR, but not all. As a results GABA inhibitory action diminishes while glutamate-mediated excitation persists, leading to an increase dopaminergic neurons firing and enhancement in responsiveness to nicotine (Mansvelder & McGhee, 2000, 2002).
Nicotine also affects the release of endogenous opioid peptides. Nicotine binding to nAChR within hypothalamus induces the release of a precursor of ß-endorphin. It is thought to be involved in mood regulation, decrease response to stress, conserve energy and relaxation (Cesselin, 1995).
As far as concerns serotonergic transmission, it has been shown that chronic nicotine exposure produces a selective decrease in the concentration of 5-HT in the hippocampus (Benwell & Balfour, 1979). The effect of this neuroadaptation is still unclear, however, considering the findings that 5-HT deficits have been implicated in depression and anxiety, it may be hypothesized that during chronic nicotine exposure and withdrawal, the decrease in serotonin function play a role in the onset of negative affective symptoms, such as depressed mood and irritability (Schwartz, 1984).
1.1.4 Nicotine effects and withdrawal
The activation of peripheral nAChRs increases noradrenaline release, with concomitant increases in heart rate, blood pressure, and respiratory rate. Centrally nicotine improves working memory functions, learning and attention; it also induces pleasure and reduces stress and anxiety. At the initial experience it can give nausea/disorientation.
After a first experience of smoking, as a result of pharmacological and non-pharmacological factors, an individual frequently elect to repeat the experience (Rose, 2006). This leads to the next stage where the prolonged exposure to smoke induce a neuroadaptation in the brain, increasing the reinforcing effects of nicotine (Soria, et al., 1996). When CNS nicotine levels ceases abruptly following smoking cessation, it produce temporary imbalances in neurological systems before compensatory mechanisms are triggered to restore homeostasis (Lowinson, 2005). This imbalance is
13
associated with unpleasant withdrawal effects such as irritability, headache, nausea, constipation or diarrhoea, falling heart rate and blood pressure, fatigue, drowsiness or insomnia, depression, increased hunger and energy, lack of concentration, anxiety, and cravings for cigarettes (Benowitz, 1988) which are powerful incentives to take up/relapse smoking again (Hughes, 1992; Hughes et al., 1984; 1991) Thus basis of nicotine addiction is a combination of positive reinforcement of mood and avoidance of withdrawal symptoms. In addition, conditioning has an important role in the development of tobacco addiction.
1.1.5 Pharmacological smoking cessation treatment
First-line pharmacological treatments of tobacco dependence recommended by clinical practice guidelines are nicotine replacement therapy (NRT), bupropion and varenicline (Lerman et al., 2007).
Nicotine replacement therapy (NRT) is the only first-line smoking cessation treatment available without prescription and has increased short-term smoking cessation rates by 50–70% (Rigotti, 2002). NRT reduces the severity of withdrawal symptoms such as anxiety, insomnia, depressed mood, and inability to concentrate (Ford and Zlabek, 2005). Smoking whilst using NRT provides a deterrent, as the high nicotine doses can produce aversive effects such as nausea, palpitations, hypotension, and altered respiration (Frishman, 2007). NRT treatments are available as a nasal spray, chewing gum or transdermal patches. However, despite initial benefits, around 95% of ex-smokers who had undergone transdermal patch NRT relapsed after a period of time (Clinical Practice Guideline Treating Tobacco Use and Dependence 2008 Update Panel, Liaisons, and Staff. U.S.A. Public Health Service report. Am J Prev Med. 2008 35:158-76.).
Bupropion is an antidepressant drug; its primary pharmacological action is thought to be noradrenergic and dopaminergic reuptake inhibition. It binds selectively to DA transporter, but its behavioural effects have often been attributed to its inhibition of noradrenaline reuptake (Balfour, 2011). It also acts as a nAChRs antagonist. Its efficacy might be explained by its antidepressant effect, indeed depression is a withdrawal symptom that reliably predict relapse among abstinent smokers (Hughes, 2007). Moreover, its antagonist-like activity on nAChR decreases the reinforcing effect of nicotine.
liability of nicotine. Indeed it produce a lesser, slower DA release then nicotine with a longer duration if action, Moreover, when varenicline is combined with nicotine, it attenuates nicotine induced DA release in nucleus accumbens (Rolleman et al., 2007). Behavioural interventions play an integral role in smoking cessation, either in conjunction with medication or alone. They employ a variety of methods to assist smokers in quitting, ranging from self-help materials to individual cognitive-behavioural therapy. These interventions teach individuals to recognize high-risk smoking situations, develop alternative coping strategies, manage stress, improve problem solving skills, as well as increase social support (Clinical Practice Guideline Treating Tobacco Use and Dependence 2008 Update Panel, Liaisons, and Staff. U.S.A. Public Health Service report. Am J Prev Med. 2008 35:158-76.).
1.2 Psychobiology of tobacco addiction
The severity of nicotine dependence (abuse liability, frequency of consumption, high rate of relapse) is similar to other drug dependence, such as opiates or cocaine. In contrast, the reinforcing properties of nicotine is subtler compared to other drug. It suggests that the reinforcing effect of nicotine is necessary but not sufficient to explain tobacco dependence (Caggiula, 2001). Furthermore several preclinical and clinical studies have underlined the importance of non-pharmacological factors, such as environmental stimuli, in maintaining smoking behaviour and promoting relapse.
1.2.1 Conditioning
A stimulus that is repeatedly and contingently paired with an unconditioned stimulus (e.g. nicotine effect) acquires a Pavlovian conditioned value (Pavlov, 1927). Thus with regular smoking within a complex individual and social context, smokers associate specific situation, mood or environmental factors with the rewarding effect of nicotine. These smoking-associated stimuli may trigger physiological, psychological and behavioural reactivity in smokers, and it is widely accepted that they can precipitate relapse in ex-addicts (Abrams, 1999; Drummond, 2000; Niaura et al., 1988). There are two classes of conditioned stimuli: proximal discrete cues that become conditioned stimuli (CS) after association to drug effects (e.g. cigarettes, lighter), and distal stimuli that are present in the environmental context (e.g. bar and people around) (Conklin et al., 2008).
15 1.2.2 Nicotine’s multiple-action
Several studies suggest that in addition to its primary reinforcing properties, nicotine has a second effect that may be important in promoting smoking behaviour. Nicotine is a cognitive enhancer drug and may enhance the salience of other reinforcers, including the CS that has acquired conditioned values by repeated pairing with nicotine effect (Caggiula et al, 2002). Nicotine activates and potentiates information processing at those brain area and pathway where reinforcement and sensory transmission are integrated into emotional, motivational and cognitive processes that control for smoking behaviour. Smoking behaviour may therefore be maintained by a “multiple-action” effect of nicotine: i) as a primary reinforcement and ii) as an enhancer of the multiple smoking/smoking-associated stimuli processing. This model may help to explain how nicotine could play a central role in initiation, maintenance and difficulty to stop smoking, despite of its mild reinforcing properties (Chiamulera, 2005).
1.2.3. Cue reactivity
Cue reactivity is the vast array of responses that are observed when addicts or ex-addicts are exposed to drug-related CS (Drummond, 2000). These responses can be i) physiological, such as decrease in heart rate and blood pressure and/or increase in skin conductance and skin temperature, ii) psychological, such as increase in craving and urge to smoke and/or mood change, iii) and also behavioural, such as cigarette-seeking and change in smoking behaviour (e.g. latency to smoke, cigarette puff volume and frequency, amount of cigarette consumed and relapse to smoking behaviour). Several factors may influences smokers’ cue reactivity: type of stimuli (e.g. distal vs. proximal) (Conklin et al, 2008), degree of nicotine dependence (Payne et al, 1996) impulsivity, genetic, comorbidity (Drummond, 2000), contextual factor drug-availability or expectation (Field & Duka, 2001).
Several brain imaging studies have revealed that brain area of the mesocorticolimbic system are specifically activated in smokers exposed to smoking-associated stimuli, and that these effects may overlap with those induced by nicotine administration. The fact that exposure to smoking cues and nicotine administration activate similar brain patterns suggests a causal relationship between nicotine effect through smoking and development/ maintenance of cue reactivity (Yalachkov et al, 2009). Cue reactivity may last in ex smokers even after years of smoking cessation, and is the main cause of relapse to smoking behaviour (Shiffman, 2009).
1.3 Nicotine-related memories
Given the importance of the learned association between stimuli and drug, that we can also call drug-associated memories, in the phenomenon of relapse to drug-seeking behaviour, it has been proposed that treatment that disrupt the drug-associated memories could act as a pro-abstinent and anti-relapse therapy (Diergaarde, 2008; Tronson &Taylor, 2007; Taylor, 2011). Therefore there is an increasing interest in investigate the phenomena of drug memories consolidation and reconsolidation.
1.3.1 Reconsolidation theory
Memories are stored after a learning experience through a process called consolidation. For more than 100 years the idea that once consolidated memories become permanently stored in the wiring of the brain has been a dogma. In the traditional consolidation theory new memory are initially in a “labile” form for a short time (short term memory-STM), after which the memory trace is fixed or “consolidated” into the physical structure of the brain (long term memory-LTM). In 1968 Lewis and colleagues observed that an electroconvulsive shock (an amnestic treatment), provided after the memories have been reactivated by its retrieval, could induce amnesia the following day. Given that amnesia was not produced in the absence of memory reactivation it has been argued that retrieval of memory induce a reactivation of the memory trace, that presumably return to a labile state, which initiated another memory process similar to that seen after learning. The processes through which memory are maintained after retrieval is called reconsolidation (Figure 3).
17
Figure 3: Two model of memory processing. (a) The traditional consolidation memory that stated that a labile short-term memory (STM) and a later, consolidated, permanent long-term memory (b) The memory model proposed by Lewis (1968). The active state (AS) and inactive state (IS) are analogous to STM and LTM, respectively. Memory after learning experience are in AS, then it enter in IS by the passage of time. Retrieval of the memory returns it to the AS (Nader, 2003).
Furthermore it has been shown that amnesia can be induced only if the amnestic treatment, such as the electroconvulsive shock, is given shortly after retrieval (Misanin et al, 1968; Schneider & Sherman; 1968). These findings suggest that retrieval induce a transient labile phase of the memory, the time during which memory trace are labile is called reconsolidation window and persist for several hours after retrieval (Duvarci & Nader, 2004; Nader et al, 2000; Sara, 2000).
In the past 10 years the study of reconsolidation have been extended to numerous
Memory traces unbound
Karim Nader
Department of Psychology, McGill University, Montreal, Quebec, Canada H3A 1B1
The idea that new memories are initially ‘labile’ and sensitive to disruption before becoming permanently stored in the wiring of the brain has been dogma for >100 years. Recently, we have revisited the hypothesis that reactivation of a consolidated memory can return it to a labile, sensitive state – in which it can be modified, strengthened, changed or even erased! The data gener-ated from some of the best-described paradigms in memory research, in conjunction with powerful neuro-biological technologies, have provided striking support for a very dynamic neurobiological basis of memory, which is beginning to overturn the old dogma.
For .100 years, generations of behavioural paradigms and technologies have been used to address questions about the mechanisms that mediate learning and memory
[1–3]. Repeatedly, evidence has been found to suggest that the properties of the memory trace change in a time-dependent manner, such that new memories are initially in a dynamic ‘labile’ form for a short time [short-term memory (STM)], after which the memory trace is ‘fixed’ or ‘consolidated’ into the physical structure of the brain [long-term memory (LTM)][4–6]. For example, electroconvul-sive shock (ECS) is effective in inducing amnesia if presented shortly after training (during STM) but not if given a few hours later (during LTM)[7]. Time-dependent effects such as these are the cornerstone of memory consolidation theory (now called cellular consolidation theory [8]). During the past 40 years, incredible efforts have been made to describe across all levels of analysis the processes that contribute to the transformation of a trace from being labile to being fixed[9,10]. Of note is the finding that the transcription factor Ca2þ -response-element-binding protein (CREB), transcription and trans-lation all seem to be universal neuronal requirements for traces to enter LTM[11–15](Fig. 1a).
Early studies on reconsolidation
In 1968, the view that memories are consolidated over time into a permanent state was challenged by Lewis and colleagues[16]. In agreement with previous studies, when ECS was given 24 h after fear conditioning it was ineffective in generating amnesia. However, if the memory was reactivated before ECS administration, amnesia was observed the following day. Given that amnesia was not produced in the absence of memory reactivation, the memory is defined as being consolidated by that time. Therefore, reactivation of a consolidated memory presum-ably returned it to a labile state, which initiated another
time-dependent memory process similar to that seen after new learning. This phenomenon is now referred to as reconsolidation[17–19]. Lewis’ study defined a paradigm for experimentally differentiating consolidation and recon-solidation: a necessary criterion if an effect is to be attri-buted to reconsolidation is that the amnesic agent must be
Fig. 1. Two models of memory processing. (a) The traditional consolidation theory, which posits a labile, short-term memory (STM) state and a later, consolidated long-term memory (LTM) state. Once fixed in LTM, the memory is posited to be permanent. Below each memory state is a list that is typically used to describe some of the properties of the two states. (b) The memory model proposed by Lewis[33]. The active state (AS) and inactive state (IS) are analogous to STM and LTM, respectively. The molecular descriptors in brackets were not part of the original model but have been inserted for comparison with (a). New memories enter a labile AS and then with time enter the IS [top red arrow, again similar to (a)]. Reactivation of memories that are in an IS returns them to the AS (bottom red arrow). Both new and reactivated memories require protein-synthesis-dependent mechanisms in order to enter the IS. Contrary to consolidation theory, which cannot explain the reconsolidation data, this model incorporates both the data from consolidation and reconsolidation experiments.
TRENDS in Neurosciences
Short-term memory (STM)
• Lasts for seconds to hours • ‘Labile’ (sensitive to disruption) • Does not require new RNA or protein synthesis
Active state (AS)
• Lasts for seconds to hours • ‘Labile’ (sensitive to disruption) (Does not require new RNA or protein synthesis)
Long-term memory (LTM)
• Lasts for days to weeks • Consolidated (insensitive to disruption)
• Does require new RNA or protein synthesis
Inactive state (IS)
• Lasts for days to weeks • Inactive (insensitive to disruption)
(Does require new RNA or protein synthesis)
(a)
(b)
Corresponding author: Karim Nader (nader@hebb.psych.mcgill.ca).
Opinion TRENDS in Neurosciences Vol.26 No.2 February 2003 65
species, including crabs, chicks, honey bees, etc. and to numerous experimental paradigms (Figure 4).
Figure 4: Example of experimental paradigms and treatment and species involved for studies that reported evidence of reconsolidation process since 2000. (Modified from Nader & Hardt, 2009).
To experimentally demonstrate reconsolidation or the role of a particular molecule in reconsolidation memories must be first consolidated, then reactivated (retrieved) contiguously with some form of manipulation. Finally, modification of the memory must be observed.
Reconsolidation is frequently studied using Pavlovian conditioning paradigms, such as fear conditioning. Training consists of pairing a neutral stimulus (conditioned stimulus-cue), such as a tone, with a reinforcing stimulus (unconditioned stimulus), such as a foot-shock. Retrieval is induced in a reactivation session, which occur at least 24 hours later and consists in presenting the conditioned stimulus in the absence of unconditioned stimulus. The manipulation (such as the administration of an amnestic drug) is applied either prior or immediately after the reactivation session. Finally at least 24 hours later the memory is tested by re-presenting the cues and measuring the unconditioned responding, in this case the freezing (measure of fear response), compared with animal
reactivation-dependent interference effects in consoli-dated episodic memory were found only when human subjects were exposed to the interfering material in the same environment in which the original learning took place. Thus, activating memory outside the spatial learn-ing context was not sufficient to induce reconsolidation. So-called boundary conditions are physiological, envi-ronmental or psychological situations in which memory that normally would reconsolidate does not. Several boundary conditions have been proposed, such as extinc-tion consolidaextinc-tion65,66,68, memory age67,68, predictability
of the reactivation stimulus74,75 and training intensity68.
However, these results remain controversial, as other studies were unable to replicate them (for extinction see
REFS 69,70; for old memories see REFS 37,77; for
predict-ability of the reactivation stimulus see REFS 36,40,58,78;
and for strength of training see REFS 37,77).
Therefore, it remains to be seen whether additional parameters moderate boundary conditions.
There is currently no universally applicable recon-solidation protocol to reliably destabilize consolidated memory, which in turn complicates establishing bound-ary conditions. If under certain conditions reconsolida-tion effects are not detected, one cannot conclude with certainty that a boundary condition has been found. For example, in contextual fear conditioning, memo-ries that were acquired with a strong training protocol of three shocks did not undergo reconsolidation if the reactivation session took 3 or 5 minutes, but reactivat-ing memory for 10 minutes triggered reconsolidation68.
If only the two shorter reactivations had been used, the absence of reconsolidation might have been taken as evidence that memories acquired with strong training do not undergo reconsolidation, implying a true bound-ary condition. This kind of parametric manipulation has not been performed for most proposed boundary con-ditions suggested by certain experimental results. It is thus unclear whether these conditions are true boundary conditions or merely situations in which it is harder than normal to induce reconsolidation.
Alternative interpretations
Reconsolidation, as discussed above, has been defined using the very standards that define consolidation. Therefore, questioning certain aspects of the reconsoli-dation hypothesis poses the same challenges for the con-solidation hypothesis. The reconcon-solidation hypothesis in its current form has come under considerable scrutiny. We now discuss some of the alternative interpretations of the data that have been proposed in the literature.
Lesion or nonspecific effects. One theory posits that the
amnesic treatment induces a lesion and thus impairs reactivated consolidated memory79(FIG. 3). This
sugges-tion explains the deficits in PR-LTM and, if the treat-ment took several hours to destroy the local tissue and produce the lesion, it might explain the intact PR-STM. However, amnesic animals can be retrained, demonstrat-ing that the targeted brain structures remain functional80.
Moreover, without memory reactivation the amnesic Table 1 | Some of the paradigms in which reconsolidation has been reported
Experimental paradigm Treatment Animal Refs
Habituation Heat shock, and DNQX (antagonist of non-NMDA-type glutamate receptor) Nematode 59 Auditory fear conditioning Protein synthesis inhibition, inhibition of kinase activity, and reconsolidation
potentiation by protein kinase A activation Rat 35,99,137
Classical fear conditioning Transient anaesthesia Medaka (a fish) 65
‘Pavlovian-like’ conditioning Protein synthesis inhibition, sensory block, mRNA synthesis inhibition and blocking
bond formation of cell-adhesion molecules Hermissenda 42
Contextual fear conditioning Protein synthesis inhibition, inducible CREB-knockout and antisense
oligodeoxynucleotides Rat and mouse 37–39
Context-signal memory NMDA receptor antagonist Crab 74
Operant conditioning RNA synthesis inhibition, and cooling Snail 40
Appetitive conditioning Protein synthesis inhibtion Honeybee 69
Conditioned taste aversion Protein synthesis inhibition Rat pups 138
Inhibitory avoidance Protein synthesis inhibition, glycoprotein synthesis inhibition and antisense
oligodeoxynucleotides Chicks and rats 67,87,139
Motor sequence learning Interference by new learning Humans 41
Incentive learning Protein synthesis inhibition Rat 140
Object recognition Zif268-deficient mouse, and inhibition of kinase activity Mouse, rat 36,141
Spatial memory Protein synthesis inhibition Mouse and rat 68,75
Memory for drug reward Inhibition of the ERK kinase MEK, Zif268-deficient knock-in mice and Zif268 antisense
oligodeoxynucleotides Rat and knock-in mouse 57,58,77
Episodic memory Interference by new learning Humans 72,76
This table lists example experimental paradigms and the treatments and species involved for studies that reported evidence of a reconsolidation process since 2000. CREB, cAMP-responsive element-binding protein; DNQX, 6,7-dinitroquinoxaline-2,3-dione; ERK,extracellular signal-regulated kinase; MEK, MAPK/ERK kinase; NMDA, N-methyl-Ż-aspartate.
R E V I E W S
228 | MARCH 2009 | VOLUME 10 www.nature.com/reviews/neuro
19 in the non-manipulated control group.
Demonstrating reconsolidation not only requires evidence of modification of a previously consolidated memory, but also evidence that in the absence of retrieval or if the amnestic manipulation is applied outside the reconsolidation window, the memory remains unmodified.
To better understand the cellular a molecular mechanisms underlying of particular focus have been the molecular cascades previously demonstrated to be important in memory consolidation and those downstream of therapeutically relevant neurotransmitter targets including β -adrenergic receptors and NMDARs (N-methyl-d-aspartate receptors). De-novo protein synthesis is required for memory reconsolidation; several animal studies have shown that injection of protein synthesis inhibitor, such as anisomycin, after retrieval of a previously consolidated memory, can disrupt the original memory. It has been shown that the immediate-early genes c-Fos and JunB are activated during, and CCAAT-enhancing binding protein-β (C/EBPβ) is required for, memory reconsolidation. The gene transcription is initiate by the activation of transcription factors such as cAMP response element-binding protein (CREB), zinc finger 268 (zif-268), ELK1 and nuclear factor kB (NF-kB). These, in turn, are activated by upstream kinase, such as extracellular-regulated kinase (ERK) and protein kinase A (PKA) (for review see Tronson & Taylor, 2007) (Figure 5).
Figure 5: Key molecular mechanisms of memory reconsolidation. Molecular signalling cascades downstream of β-adrenergic receptors (β-AR) and NMDARs (N-methyl-d-aspartate receptors) have been shown to be implicated in reconsolidation. Small GTPases such as Ras, Raf and Rap activated by Ca2+ influx activate the extracellular signal-regulated kinase pathway (ERK). Protein kinase A (PKA) is activated by cyclic AMP (cAMP) and acts directly, or indirectly through ERK and ribosomal protein S6 kinase (RSK), to activate transcription factors including cAMP response element-binding protein (CREB), zinc finger 268 (ZIF268) and ELK1, which then initiate gene transcription. The immediate-early genes c-Fos and JunB are activated during, and CCAAT-enhancing binding protein-β (C/EBPβ) is required for, memory reconsolidation
(image taken from Tronson &Taylor, 2007).
From an evolutionary perspective, it has been argued that reconsolidation may serve as an adaptive update mechanism allowing for new information, available at the time of retrieval, to be integrated into the initial memory representation (Alberini, 2005; Hupbach et al, 2007; Monfils et al, 2009; Nader, 2002). Other authors proposed that
SRE AP1 CRE TATA
ELK1 SRF CREB CBP SRF CREB PKA B-Raf Rap1 MEK ATP cAMP ERK Ras Raf1 P P P P P P P NMDAR β-AR RSK Nucleus Intracellular Extracellular Gene transcription protein synthesis ZIF268 ZIF268 mRNA ZIF268 ZIF268 Gene transcription protein synthesis IEGs: c-Fos JunB, etc. C/EBPβ CaMKK CaMKIV Gαs βγ Adenylylcyclase Ca2+ CaM C/EBPβ Glutamate Noradrenaline
conditioned place preference38,48.
PKA is also required for reconsolidation of auditory fear memories. Inhibition of PKA in the BLA by infu-sions of Rp-cAMPS, a PKA inhibitor, after memory retrieval disrupts auditory fear memories6(FIG. 3) or
conditioned taste aversion memories49. Moreover,
post-reactivation activation of PKA by injections of the PKA activator 6-BNZ-cAMP in the BLA enhances reconsoli-dation of an auditory fear memory6. Unlike its
involve-ment in memory reconsolidation, amygdalar PKA does not seem to be involved in extinction of fear, indicat-ing differential molecular or anatomical mechanisms in these two co-occurring processes6. However, PKA
is not always involved in reconsolidation in every
spe-cies; a recent study showed that retrieval of a memory shortly (6 hours) — but not 24 hours — after train-ing triggers PKA-dependent reconsolidation50. At
both times reconsolidation is PSI-dependent. This study extends previous models that have shown that older memories are more resistant to reconsolidation to suggest that, in addition, different processes are involved in reconsolidation of older than newer mem-ories. Whether such differential involvement of PKA in memories at different times after training is true in mammalian models, or other types of memory, is as yet unknown.
Immediate-early genes. Molecular events in
reconsolida-tion have also been examined by imaging cellular
activ-Figure 2 | Key molecular mechanisms of memory reconsolidation. Many individual molecules have been identified as being required for memory reconsolidation; however, few papers have put together schematic models for the pathways involved. This figure integrates findings from several studies. Of particular focus have been the molecular cascades previously demonstrated to be important in memory consolidation and those downstream of therapeutically relevant neurotransmitter targets including β-adrenergic receptors (β-AR)70,71,87–90 and NMDARs9,60,91–93 (N-methyl-d-aspartate
receptors). Molecular signalling cascades downstream of these receptors have been implicated in reconsolidation. Small GTPases such as Ras, Raf and Rap activated by Ca2+ influx activate the extracellular signal-regulated kinase pathway
(ERK)38,46–48,94. Protein kinase A (PKA)6,49,50 is activated by cyclic AMP (cAMP) and acts directly, or indirectly through ERK and
ribosomal protein S6 kinase (RSK), to activate transcription factors including cAMP response element-binding protein (CREB)15,37,38, zinc finger 268 (ZIF268) (REFS 41–45,51,52) and ELK1 (REF. 38), which then initiate gene transcription. The
immediate-early genes c-Fos and JunB37,38,53–55 are activated during, and CCAAT-enhancing binding protein-β(C/EBPβ)30,34
is required for, memory reconsolidation. Integrating all the available data aims to identify logical pathways to examine next. For example, a role for the calcium/calmodulin (CaM)–CaM-dependent protein kinase kinase (CaMKK)–CaMKIV cascade in memory reconsolidation might be inferred from NMDAR activity; however, the involvement of this pathway has not directly been examined. AP1, activator protein complex 1 (a complex of c-Fos and c-JUN); CBP, CREB binding protein; MEK, mitogen-activated protein kinase/ERK kinase; SRE, serine response element; SRF, serum response factor; TATA, box required for transcription. Figure modified, with permission, from Nature Reviews NeuroscienceREF. 76 (2001) Macmillan Publishers Ltd.
21
reconsolidation might serve to strengthen memory (Inda et al, 2011; Lee, 2009; Sara, 2000).
As stated above, it has been shown in several animal studies that memory could also be disrupted acting on the molecular mechanisms underlying reconsolidation (for review see Tronson and Taylor, 2000; Nader et al, 2000; Soeter & Kindt, 2011). This offers a potential for the treatment of psychiatric disorders characterized by strong pathogenic memories, such as post-traumatic stress disorders (PTSD), phobias and also drug addiction (Centonze et al, 2005).
1.3.2 Reconsolidation as a potential target in drug addiction treatment
Drug addiction is a chronic disorder characterized by a high rate of relapse to drug use among abstinent. One of the main causes of relapse is the exposure to the CS that are associated to drug effect in a Pavlovian manner and influence drug-seeking behaviour and relapse through the memory they evoke (Milton and Everitt, 2010).
Therefore molecular and neuroanatomical processes involved in the reconsolidation of drugs-associated memories have been proposed as novel targets for the treatment of vulnerability to CS in drug addicts (Tronson & Taylor, 2007; Diergaarde et al., 2008; Taylor et al., 2009; Milton & Everitt, 2010). Mechanistic studies identified receptors, signalling molecules and transcription factors underlying drugs-associated memory reconsolidation (Sadler et al., 2007; Brown et al., 2007; Fricks-Gleason & Marshall, 2008; Itzhak, 2008; Lee & Everitt, 2008; Milton et al., 2008a, 2008b; Fuchs et al., 2009; Ramirez et al., 2009; Sanchez et al., 2010; Théberge et al., 2010; Wu et al., 2011;). These studies have been focused mostly upon two neurotransmitters receptors, known to be involved in the reconsolidation of emotional memories: NMDA subtype of glutamate receptor and β-adrenergic receptor.
It has been shown that NMDAR antagonist, such as MK-801 or D(-)-(2R)-amino-5-phosphonovaleric acid (D-APV), given shortly after retrieval, may inhibit the reconsolidation of drug-associated memory in different Pavlovian conditioning paradigms in rats, such as conditioned place preference produced by cocaine (Kelley et al, 2007), amphetamine (Sandler et al, 2007; Sakurai et al, 2007), and morphine (Zhai et al, 2008); cue-induced reinstatement of alcohol seeking (von der Goltz at al, 2009); and the acquisition of a new instrumental response for a CS previously paired with cocaine (Milton et al, 2008). It has been proposed that a reduction in the expression of the immediate-early –gene zif268 is linked to disruption of memory reconsolidation. Indeed
Milton and colleagues found that administration of the NMDAR antagonist D-APV into the basolateral amygdala before a memory reactivation disrupt the reconsolidation of cocaine-associated memory in rats trained to cocaine self-administration and this effect is associated with a reduction in the expression of zif268. Also Lee (2005) showed that an infusion of the zif268 antisense oligodeoxynuclotides (ASO) into the basolateral amygdala contingently upon retrieval of cocaine-associated memory could disrupt the conditioned reinforcing value of the CS. However it has not yet been investigated through which signalling cascade (e.g. ERK activation or protein kinase A) expression of zif268 is linked to NMDAR. On the other hand, in rats trained to self-administer cocaine, systemic administration of MK-801 contingently upon retrieval, showed no effect on subsequent cocaine-primed reinstatement of cocaine-seeking behaviour (Brown et al, 2008).
The first evidence of the role of β-adrenergic receptor in the reconsolidation of appetitive memory had been provided by Diegaarde and colleagues in 2006: in their work they showed that the administration of propranolol, an antagonist of β -adrenergic receptor, contingently upon retrieval, could reduce the context-induced reinstatement of sucrose seeking behaviour in rats trained to sucrose self-administration. Subsequently Milton and colleagues (2008) showed that the administration of propranolol in rats trained to self-administer cocaine resulted in a retrieval-dependent impairment in the acquisition of a new response for cocaine-conditioned reinforcement, suggesting that reconsolidation of cocaine-associated memories has been disrupted. Moreover it has been shown that propranolol, administered upon retrieval, could disrupt place preference conditioned by cocaine (Bernardi et al, 2006) morphine (Robinson & Franklin, 2007a). However propranolol, given at retrieval, failed in reducing cue-induced reinstatement of cocaine seeking behaviour, following forced abstinence, in rats trained to cocaine self-administration (Milton & Everitt 2010).
Unfortunately the drugs used to disrupt memory reconsolidation in animal cannot be used in human, since they are toxic. The only drug that can be administered in human is propranolol, even if it could lead to side effect, such as orthostatic hypothension, decreased libido, bronchial adverse effects in patients affected by respiratory disease. Since there is an unmet need of novel interventions to be integrated in the current therapy, a new strategy based on targeting reconsolidation should guarantee efficacy and tolerability.
23
1.3.3 Extinction therapy
One of the most used cognitive-behavioural therapies for the treatment of anxiety disorder and to prevent the relapse in ex-drug addicts is extinction (also called Cue-Exposure). Extinction consists in the repeated presentation of previously CS in the absence of unconditioned stimulus. It is widely accepted that extinction is a new learning process, through which CS become associated to no US, leading to a decrease of the conditioned response. Extinction does not erase the original associative (CS-US) memory but instead generate a competitive inhibitory memory capable of temporally suppressing the expression of the original conditioned response (Pavlov, 1927; Quirk & Mueller 2008; Pape and Pare, 2010). Indeed original memory and extinguished response may re-emerge under three general condition: i) reinstatement, when US is presented unexpectedly (Pavlov 1927; Rescorla and Heth 1975; Westbrook et al. 2002, de Wit & Steward, 1981, Shaham et al, 1994), ii) renewal, when CS is presented outside the extinction context (Bouton and Bolles 1979) and iii) spontaneous recovery, when a certain amount of time has passed.
Several efforts have been dedicated to enhance the efficacy of extinction. It is widely acknowledged that glutamatergic NMDA receptor is directly involved in the formation of new learning and memories (Walker & Davis, 2002), and in early rodent studies it has been shown that a NMDA partial agonist D-cycloserine (DCS) may facilitate extinction (Liu et al, 2009; Falls et al, 1992; for review see Ganasen et al, 2010). Other compounds may be useful for strengthening or accelerating extinction, as suggested by recent rodent studies. These include fibroblast growth factor, methylene blue, endocannabinoids and yohimbine, N-acetylcysteine (Chatwal et al, 2009; Gonzales- Lima & Bruchey, 2004; Graham and Richarson, 2010; Morris & Bouton, 2007; Zhou & Kalivas, 2008).
1.3.4 Extinction and reconsolidation interaction.
Thus current research for the treatment of neuropsychiatric disorders based upon maladaptive memories, including drug addiction, is focused on the facilitation of extinction and on the disruption of maladaptive memory reconsolidation.
Unfortunately the compounds used in animal studies to block the reconsolidation, such as protein synthesis inhibitor (eg., anisomycin) and NMDAR antagonist (such as MK801) cannot be used in human, given their toxicity and bioavailability constraints. Moreover the efficacy of pharmacological improvement of extinction therapy is
controversial (Marissen et al. 2007).
An important step forward came from Monfils and colleagues in 2009. Capitalizing on reconsolidation as an update mechanism that allow for new information available at the time of retrieval to be integrated in the original memory trace, they hypothesized that proving no-fearful information through extinction training during the labile phase induced by retrieval, would lead to a modification of the original memory, reinterpreting CS as safe and therefore would prevent the CS-induced return of fear. They trained rats to Pavlovian fear conditioning; 24 hours later fear memory was reactivated by a single presentation of CS; 10 minutes, 1 hour, 6 or 24 hours later animal underwent an extinction session while CS was repeatedly presented in the absence of US (non fearful information). The day after they tested the return of fear under reinstatement or renewal conditions and one month later they tested the spontaneous recovery of fear. Results showed that extinction, only when applied within the reconsolidation window (10 minutes or 1 hour after retrieval) but not when applied outside the reconsolidation window (6 or 24 hours after retrieval), interfered with fear memory update and prevented fear conditioned responses such as renewal, reinstatement and spontaneous recovery. Additionally rats that received extinction without retrieval of the CS showed re-emergence of fear under renewal, reinstatement and also spontaneous recovery.
These findings were supported by the work of Schiller and colleagues in 2010. Using human electrodermal fear conditioning model they demonstrated that extinction, applied 10 minutes after retrieval (single CS presentation) of a fear memory, prevented the spontaneous recovery and reinstatement of fear response. Conversely extinction applied 6 hours after retrieval (outside the reconsolidation window) had no effect.
These two studies suggest that a new learning may interfere with memory reconsolidation of the original memory, this notion has received support also from other studies targeting episodic, motor and declarative memories both in human and in laboratory animals. Boccia and co-workers in 2005 showed the exposure to a new learning task could affect the memory reconsolidation in an inhibitory avoidance task in mice. Forcato and colleagues in 2007, 2009 and 2010 demonstrated that new information provided within the reconsolidation window may modify the original declarative memories in human. There are evidences that also episodic memories could be selectively impaired following retrieval (Hupbach et al, 2007; Strange et al, 2010). Interestingly, Flavell et al. (2011) have recently shown extinction given in conjunction
25
to retrieval was able to block the reconsolidation of appetitive memory in rats. They have used the paradigm of acquisition of new response for stimuli previously paired to sucrose. In this experimental paradigm rats are initially trained to self-administer sucrose by an instrumental response (e.g. nose-poke), and each sucrose administration is paired with the presentation of a CS, such as a light or a tone. In a second phase, rats are required to acquire a new instrumental response (e.g. lever press) to receive a CS presentation. Therefore the new instrumental behaviour is supported by the conditioned reinforcing properties of the CS. In this study Flavell et al. showed that CS-extinction applied after the retrieval of the CS, inhibited the acquisition of new response in rats trained to self-administer sucrose. This effect was retrieval-dependent since no effect was observed when extinction was applied without previous retrieval of the CS. They hypothesized that extinction applied within the vulnerable phase of the retrieved memory, was interfering with their reconsolidation. However they also pointed out that it was equally plausible that prior retrieval of the memory might facilitate extinction and therefore potentiate its effect, in a similar manner to pharmacological enhancement of extinction. To disentangle this account in a new groups of rats they substitute retrieval with the administration of D-cycloserine (DCS), an NMDA receptor partial agonist, well know to enhance extinction of memory (see paragraph 1.3.3.). Results showed that when rats were injected with DCS, instead of being retrieved, before the administration of extinction there was no effect on subsequent acquisition of new response with conditioned reinforcer. Therefore they argued that the observed post-retrieval extinction effect was due to the interference with reconsolidation of sucrose-related memories. Furthermore Flavell et al investigated the effect on retrieval-extinction procedure on reconsolidation of contextual fear memory. They showed that extinction, only when applied in combination to retrieval, prevented the return of fear in the subsequent test. Further evidence that the effect of extinction was retrieval dependent came from the fact that injection nimodipine, a blocker of L-type voltage-gated calcium channel (LVGCC) known to block the destabilization of memory at retrieval, immediately after retrieval impaired the effect of retrieval-extinction in preventing the return of fear. This result provides further evidence that extinction applied after retrieval inhibits the re-expression of the original memory by the disruption of memory reconsolidation. On the other Flavell and colleagues have also showed that the combination of memory retrieval and extinction did not prevent the return of fear, using the auditory fear conditioning paradigm. However they highlight that some methodological issues might explain the
contrasting results with the original finding of Monfils et al (such as training length). The molecular mechanism underlying the effect of post-retrieval extinction has been investigated by Clem & Huganir (2010), pairing fear conditioning paradigm and elettrophysiology assay. They trained animal in fear conditioning paradigm, the day after training memory was retrieved 30 minutes before extinction, and renewal and spontaneous tests performed the day after and also 5 days after retrieval-extinction. They observed that, compared to the no-retrieved groups, retrieval-extinction procedure inhibited the return of fear. Subsequently a groups of animal were injected with 1-aminoindan-1,5-dicarboxylic acid (AIDA), a competitive antagonist of AMPA receptor mGluR1, 1 hour before retrieval. Post-retrieval extinction effect in preventing the return of fear was inhibited by the previous administration of AIDA. Thus, they argued that effect of extinction upon retrieval required the mGluR1 activation. In further electrophysiological studies they observed a significant decrease of AMPA receptors – mediated transmission in the retrieved group compared to the no-retrieval group. This decrease was accompanied by the selective removal of synaptic calcium-permeable AMPA (AMPAr) receptors in the lateral amygdala. Moreover the stability of CP-AMPAr was regulated by the activation of mGluR1. Considering post-retrieval extinction effect as a reconsolidation update author suggest that mGluR1 activation is required to update memories, and that mGluR1 could be a potential drug target for preventing the return of fear.
Other studies have shown that the retrieval-extinction paradigm used by Monfils and Schiller was ineffective in preventing the return of fear in fear conditioning paradigm both in human and in laboratory animals. This type of procedure allow to isolate the acquired conditioned Pavlovian conditioned reinforcing properties if CS, from the instrumental component of the conditioning (see below, operant and Pavlovian conditioning).
First evidence came from a preclinical study by Chan and colleagues (2010): they use the same procedure described by Monfils et al (2009) to study the effect of a single CS presentation (retrieval) on the extinction and recovery of conditioned fear response via renewal and reinstatement in fear conditioning paradigm. Conversely to Monfils et al, they found that exposure to retrieval prior to extinction increased responding to that retrieved CS on subsequent test for renewal and reinstatement. The retrieval-extinction procedure has been also tested on remote fear memory (29 days old) in a mouse model of PTSD (Siegmund and Wotjak, 2007a), that compared to that used by Monfils et al,
27
take in consideration also the non associative component of fear memory (i.e. sensitization process that increase the animal response to harmless stimuli independently from the CS-US association) (Costanzi et, 2011). The main result of this study is that extinction when applied after retrieval of remote fear memory did not persistently attenuate the expression of fear. In a recent paper Pèrez-Cuesta (2009) investigated the effect of retrieval-extinction procedure in memory model of the crab Chasmagnathus (Maldonado, 2002). They trained crabs using the context-signal memory paradigm, and 24 hours later they exposed the crabs to the training context for 15 minutes (short exposure induced retrieval of the conditioned context) and 15 minutes later they exposed the crabs to the same context for an additional 2 hours (long exposure induced extinction). The day after the crab were tested for CS-US memory (test 1), and, if no memory was found, the test was replicated 24 hours later (test 2) to distinguish reconsolidation impairment (supposed to be permanent) and extinction (supposed to be transient). On test 1 no memory recovery has been observed in crabs that receive the retrieval-extinction treatment, on the contrary on test 2 a re-emergence of memory have been noticed. These data suggest that extinction applied after retrieval does not update the original memory trace.
The effect of the combination of retrieval and extinction has been also investigated using the paradigm of morphine-induced conditioned place preference (CPP) (Ma et al. 2011). They showed that repeated retrieval-extinction procedure (across 10 days) suppressed the reinstatement and spontaneous recovery of extinguished CPP. On the other hand no effect was observed when extinction was applied without prior retrieval of the memory. However recovery of the CPP was found in a reinstatement test performed 4 week after the last extinction session. The latter finding suggests that memory trace was not been erased by post-retrieval extinction. It can be hypothesized that extinction applied after retrieval did not affect the reconsolidation of memory under the conditions used in the paper by Ma et al.; otherwise, as suggested by the authors, that reconsolidation blockade did not lead to the erasure of the memory that can re-emerge by the passage of time.
As far as concern the human studies, in 2011 Soeter & Kind pointed out that the electrodermal conditioning used by Schiller and colleagues seems to primarily reflect only the cognitive level (declarative memory) of contingency learning (CS-US association), whereas human strartle potentiation is considered to be a reliable and specific index of fear. In a within-subject (Soeter & Kindt, 2011) and in a
